Getulio Cesar De Leon Fernandez
Empirical economist working where economics, mathematics, statistics and computer science meet
Economist and data professional who turns messy, large-scale data into optimal decisions. I bring a strong foundation in economic theory together with hands-on technical skills - Python, R, SQL, VBA, Power Query and Power BI - to automate workflows, model complex systems, and replace legacy processes with scalable solutions.
Read more
My background is in national accounts and macroeconomic analysis (GDP under the SNA 2008 framework), but my real interest is broader: building data pipelines, dashboards, and analytical tools that make information actually usable. I care about the full path from raw data to insight - automation, reproducibility, and clarity.
Outside of work, my personal projects explore agentic AI and LLM-driven automation - prompt engineering and agentic workflows to automate research and analysis end to end. It is where I experiment with the newest tools before they become standard.
What keeps me curious is the intersection of economics, computation, and philosophy: not just how to process data, but what it measures, what it hides, and how economic systems actually work. I am as drawn to the questions behind the numbers as to the tools that answer them, and I see that mix of theory and engineering as my real edge.
Currently pursuing a Master's in Big Data Analysis and Visualization (UNIR).
Areas
Empirical and quantitative economics, national accounts (SNA 2008), macroeconomic measurement
Applied econometrics, panel data, inferential statistics, mathematical modeling
Data visualization, dashboards, Power BI, exploratory analysis
Python, R, SQL, data pipelines, dashboards, big data
Agentic workflows, LLM-driven automation, prompt engineering
Philosophy of measurement, labour theory of value (Cockshott, Marx, Shaikh), history of economic thought
Projects
big-data-msc-portfolio
A vault - a collection of techniques and code from the Big Data MSc: MongoDB and Neo4j pipelines, supervised ML (scikit-learn, Keras), inferential statistics, and a Spark/HDFS/Kafka streaming exercise. Each module is self-contained with its own bilingual README.
finanzas-bpd
An end-to-end pipeline that turns bank notification emails into a categorized, self-hosted budget: keyword rules plus an LLM for the ambiguous cases, Apple Pay merchant matching, and a sync into a self-hosted Firefly III instance.
nominas-sector-publico-rd
Pipeline and Streamlit dashboard for Dominican public-sector payroll, built from the monthly transparency-portal PDFs. Extracts, cleans, and normalizes records to compare compensation structure and career trajectories across institutions.
obsidian-rag-brain
RAG and maintenance scripts for a personal knowledge vault: hybrid search, incremental embeddings, and a guardrailed nightly consolidation job that maintains links and graph structure without silently rewriting notes.
claude-code-skills
Eleven Claude Code skills for research, writing, and personal knowledge management: multi-source deep research, STORM-method cited reports, book-to-skill conversion, and more.
Static HTML, no build · self-hosted fonts · source on GitHub